F Le Pont, L Letrilliart, V Massari, et al and attempted suicide in : results of a general practice sentinel network, 1999–2001

Françoise Le Pont, Laurent Letrilliart, Véronique Massari, Yves Dorléans, Guy Thomas and Antoine Flahault

SUMMARY Introduction The continuous surveillance of suicide and attempted suicide cases RANCE has a high suicide incidence rate (20.0 per was added to the tasks of the French Sentinel Network of General Practitioners (GPs) in 1999. In 2001, 9700 were estimated F100 000 inhabitants in 1998), a value 2.2 times higher 1 to have occurred and an estimated 61 500 attempted suicide cases than that for the United Kingdom. However, as in other were diagnosed by GPs, representing approximately 40% of cases countries, few data on the incidence or management of nationwide. The majority of suicide and attempted suicide cases attempted suicide are available, especially in the primary involved women (67%) and 43% of all cases involved patients aged care setting.2 Continuous surveillance of suicide was there- 25–44 years. The fatality rate increased with age. About 80% of GPs fore introduced through the French Sentinel Network of complied with the current recommendation to refer patients who General Practitioners (GPs) in 1999.3 This network, created had attempted suicide to hospital. in 1984, is made up of about 1200 unpaid volunteer GPs Keywords: family practice; sentinel surveillance; suicide; suicide, distributed throughout metropolitan France. About 500 of attempted. these GPs participate on a weekly basis in the surveillance of 12 health indicators. Every week, GPs report on the number of cases using an online system. Computerised alert routines, which check for missing or discordant data, include phone calls to the GPs concerned. A detailed description of the network can be found on the Sentinel Network website (http://rhone.b3e.jussieu.fr/senti).4 We defined suicide and attempted suicide as any fatal or non-fatal act in which an individual deliberately causes self- injury (for example, self-mutilation, poisoning, jumping from high places, hanging, gunshots, asphyxiation) or ingests a medicinal product in excess of the generally recognised therapeutic dose.

Method In cases of suicide or attemped suicide, physicians report- ed on the patient’s age, sex, history of previous attempts, method used, vital outcome (whether the patient survived or died), hospital referral, and subsequent ambulatory follow-up. To obtain national incidence rates, we multiplied the mean number of cases per GP (standardised accord- ing to GP participation and geographical distribution) by the total number of GPs in France, and then divided the result by the French population. Calculation of the 95% confidence interval was based on the assumption that the number of reported cases follows Poisson’s distribution. F Le Pont, PhD, researcher; L Letrilliart, MD, PhD, coordinator of the sen- tinel network; V Massari, PhD, researcher; Y Dorléans, BSc, Fatality rate was calculated as the number of fatal suicides database manager of the sentinel network; G Thomas, MD, PhD, divided by the combined total of suicides and attempted professor of psychiatry and researcher; A Flahault, MD, PhD, professor of public health, Inserm U444, Faculté de Médecine Saint suicides with known vital outcome. Statistical comparisons Antoine, Paris, France. were performed using the χ2 test. Address for correspondence Results Françoise Le Pont, Faculté de Médecine Saint Antoine, Paris, France. E- mail: [email protected] Suicide and attempted suicide Submitted: 28 August 2002; Editor’s response: 15 August 2003; Between 1999 and 2001, 474 cases of suicide and final acceptance: 22 November 2003. attempted suicide were recorded (75 suicides, 328 ©British Journal of General Practice, 2004, 54, 282-284. attempted suicides, and 71 cases with unknown vital out- come). Cases with unknown vital outcome were either

282 British Journal of General Practice, April 2004 Brief reports

Table 1: Methods of attempted suicide and suicide among men and women (1999–2001).a HOW THIS FITS IN Attempted Attempted What do we know? suicide suicide Suicide Statistics on suicide acts are available Suicide in men in women in men through mortality certificates in France. methods (n = 81) % (n = 247) % (n = 61) % However, data on suicide attempts are lacking, especially in a primary care setting. Drug overdose 77 82 5 Other toxic What does this paper add? substance 7 3 0 Hanging 5 1 39 Data collected through the French Sentinel Network on suicide Drowning 0 0 8 and attempted suicide cases provide information on Firearms 1 1 23 incidences, patients’ characteristics, and the management of Jumping from a suicide attempts. high place 3 2 10 Self-mutilation 5 8 2 Otherb 2313 hospitalised (69%) or had unknown hospitalisation status aData are not presented for suicide in women owing to the small b (31%) and were not followed up by the GPs; they were sample size (n = 14). Subjects who used multiple methods are classified in the ‘other’ category. excluded from the analysis when studying attempted suicides and suicides separately. Assuming that all unknown vital outcome cases were attempted suicide, the than among those under 45 years (42%) (P = 0.002). The number of cases of attempted suicide diagnosed in 2001 fatality rate of medication overdose was 3%. The absence by French GPs was estimated to be 61 500 (95% confi- of previous suicide attempts (67% of suicide cases) was dence interval [CI]= 50 300 to 72 700), yielding a yearly associated with higher fatality than in cases in which incidence of 105 per 100 000 inhabitants. We estimate the suicide had been attempted previously, both overall (22% number of cases of suicide to be 9700 (95% CI = 8000 to versus 10%, respectively, P<0.05) and in the subgroup of 11 500). Only 20% of patients were younger than 25 years patients older than 44 years (38% versus 14%, P<0.05). of age, 43% were aged 25–44 years, and 37% were older than 44 years. Women represented 67% of patients Discussion overall, 19% of those who committed suicide and 75% of those who attempted suicide. The estimate of 9700 cases of suicide is in accordance with the 11 000 cases obtained from death certificates.1 In keeping with French national mortality data, hanging and Attempted suicide firearms were respectively the first and second most In twenty-three per cent of attempted suicide cases, common methods used for suicide.5 The age and sex patients were under 25 years old, whereas 43% were aged distributions were generally in keeping with surveys 25–44 years, 23% aged 45–59 years, and 11% were older conducted in other countries: patients who unsuccessfully than 59 years. The age distribution of attempted suicide are more frequently younger and female patients did not vary with sex. Medication overdose was the compared to those who successfully commit suicide.2,3 most frequently recorded method (Table 1). Hospitalisation However, several differences were observed in terms of the and ambulatory follow-up data were available in 93% of methods used and the history of previous attempts. In attempted suicide cases; 80% of patients were referred to a France, as in Belgium (33%),2 hanging was the most general hospital (with subsequent ambulatory follow-up in frequently reported method, and firearm use in France was 14%, and without in 66%), 19% had ambulatory follow-up less common than in the United States (54%)6 but more only, and 1% had no follow-up. common than in Belgium (10%).2 The rate of previous attempted suicide was found to be much higher than in Suicide Belgium (55% for France versus 30% for Belguim).2 In six per cent of suicide cases, patients were under With an annual estimate of 155 000 attempted suicides 7 25 years of age, whereas 37% were aged 25–44 years, that lead to health care demand, our data suggest that, in 20% were 45–59 years, and 37% were older than 59 years. France, GPs are the first health professionals involved in Hanging was the most frequently recorded method (36%), 40% of cases. Most GPs followed recommendations to followed by firearm use (21%). The fatality rate was much refer patients attempting suicide to hospital for psychiatric 8 higher among men than women (43% versus 5%, respect- evaluation and multidisciplinary care. ively, P<0.001) and increased with age (3% at 15–24 years, 16% at 25–44 years, 17% at 45–59 years, and 44% at over References 59 years; P<0.001). The choice of method was not the only 1. Self directed violence. In: World report on violence and health. Krug EG, Dahlberg LL, Mercy JA, et al (eds). Geneva: World explanatory factor, as the age effect persisted in the subset Health Organization, 2002: 185-212. of patients using highly lethal methods (hanging, firearms). 2. Van Casteren V, Van der Veken J, Tafforeau J, Van Oyen H. Suicide and attempted suicide reported by general practitioners The case fatality rate for the highly lethel methods sub- in Belgium, 1990–1991. Acta Psychiatr Scand 1993; 87: 451-455. group was still higher among patients over 44 years (76%) 3. Valleron A-J, Garnerin P. Computerised surveillance of

British Journal of General Practice, April 2004 283 F Le Pont, L Letrilliart, V Massari, et al

communicable diseases in France. Commun Dis Rep CDR Rev 1993; 3: 82-87. 4. Le réseau sentinelles [The sentinel network]. Who are we? http://rhone.b3e.jussieu.fr/senti/en/docs/plaquette.pdf (accessed 17 Feb 2004). 5. Institut National de la Santé et de la Recherche Médicale [National Institute of Health and Medical Research]. SC8 — Service d’information sur les causes médicales de décès [SC8 — Information service for the medical causes of death] http://sc8.vesinet.inserm.fr:1080/ (accessed 17 Feb 2004). 6. Spicer RS, Miller TR. Suicide acts in 8 states: incidence and case fatality rates by demographics and method. Am J Public Health 2000; 90: 1885-1891. 7. Badeayan G, Parayre C. Suicides et tentatives de suicides en France, une tentative de cadrage statistique [Suicides and attempted suicides in France, a tentative statistical framework]. Etudes et résultats. 2001; 109. 8. Isacsson G, Rich CL. Management of patients who deliberately harm themselves. BMJ 2001; 322: 213-215. Acknowledgements We thank the sentinel GPs who participated in the network.

284 British Journal of General Practice, April 2004